Transplant glomerulopathy (TG) is a histologically distinct entity of kidney allografts characterized by layering of capillary basement membranes, usually identified by light microscopy as a duplication of the glomerular basement membrane (1). Transplant glomerulopathy is present in approximately 4% of surveillance biopsies taken 1 year after transplantation. Even when diagnosed early, TG is progressive (2) and frequently leads to allograft failure.
Transplant glomerulopathy is more than a glomerular process, and in most cases, the disease involves the entire renal capillary network (3, 4). Transplant glomerulopathy is associated with the presence of antigraft antibodies that in most cases, but not all, are directed against human leukocyte antigen (HLA) molecules (5, 6). Recent studies further clarified these associations and postulated that specifically anti-HLA class II (HLA-II) antibodies are responsible for TG (2). These latter studies are based on the use of newer, sensitive, and solid phase technologies for the detection of anti-HLA antibodies. Using these technologies, it is becoming evident that a substantial percent of patients thought to be recipients of “conventional” transplants are in fact sensitized against HLAs before transplantation. The risk of TG associated with anti-HLA-II antibodies is variable. Thus, among antibody-positive patients, 26% develop TG before transplantation, but 74% do not (2).
The goal of this study was to further define the risk of TG, searching for additional anti-HLA antibody characteristics that may relate to the risk of developing TG and with graft survival after developing TG. Specifically, we were interested in assessing the anti-HLA antibody level and specificity in pretransplant serum; thus, we could define the risk and prognosis at the time of transplantation. These studies were performed in a patient cohort that had been previously reported (2). Reanalysis of this cohort first showed that with a longer follow-up time, the incidence of TG has continued to increase in patients with pretransplant anti-HLA-II antibodies, as predicted by previous analyses (2). Furthermore, these analyses identified the antibody level and its reactivity against donor HLAs (donor-specific activity [DSA]) as independent variables related to the risk of TG. These analyses also identified new variables related to the survival of grafts with TG. Among these, the presence of C4d in peritubular capillaries (PTCs) relates, independent from other prognostic factors, to the graft loss in TG. These results represent one more step in identifying the patients at risk for TG and for a poor transplantation outcome.
Patient Selection and Study Design
This study was conducted using a protocol approved by the Institutional Review Board of the Mayo Clinic. We retrospectively studied 598 recipients of kidney transplants at our institution (both living donor and deceased donor recipients) transplanted from January 2000 to January 2005 who had a negative T-cell anti-human globulin (AHG)-enhanced complement-dependent cytotoxicity crossmatch against their donor at the time of transplantation. B-cell crossmatch data of these patients were not available at the time of transplantation. These patients were transplanted during an era when more sensitive assays to determine anti-HLA antibody levels (T and B flow cytometric crossmatches or single-antigen flow bead assays) were not available at our institution. Thus, with the exception of high levels of anti-HLA-I antibodies that were excluded using the T-cell AHG-CDC crossmatch, many of these patients might have been transplanted with a variety of anti-HLA class I and class II antibodies with a wide spectrum of levels. Until recently, these patients were considered in this program and in most transplantation programs as “conventional” transplant recipients. Using the serum collected before transplantation, we retrospectively determined the presence of anti-HLA antibodies and estimated its level using the single-antigen flow bead assay (see later). This design allowed us to assess the impact of antibody and other risk factors for the development of TG and graft loss in an essentially blinded fashion, that is, no patient was treated differently based on their anti-HLA antibody status at the time of transplantation.
Clinical and laboratory data from donors and recipients were extracted from electronic databases and from the patients’ medical records. The characteristics of this patient population are shown in Table 1. The immunosuppression protocols used in these patients have been previously described (2). In brief, 92% of patients received induction immunosuppression (78% of cases received thymoglobulin; 13% received anti-CD25 receptor antibodies; and 1% received OKT3). Maintenance immunosuppression consisted of prednisone, tacrolimus, and mycophenolate mofetil in 81% of patients. Cyclosporine was used in 9% of patients and sirolimus in 9% instead of tacrolimus.
Characterization of Anti-Human Leukocyte Antigen Antibodies
Measurements of anti-HLA antibodies were performed in all patients using pretransplant final crossmatch sera that had been stored at −70°C. These sera were analyzed first using solid-phase antibody assays containing multiple HLAs (multi-antigen synthetic flow bead, LABScreen I and II, One Lambda Inc., Canoga Park, CA). Sera found to have anti-HLA antibodies were then tested using single HLA-coated flow beads (SAFB) (LABscreen Single Antigen, One Lambda, Inc.) to determine and confirm the antibody’s specificity. In these solid-phase assays, antibodies are detected by measuring the fluorescence intensity on each HLA-coated bead and the fluorescence intensity, in arbitrary units, expressed as the normalized value (NV) calculated as (fluorescence of beads coated with HLA and incubated with patient’s serum)−(fluorescence of beads without HLA incubated with patient serum)−(fluorescence of beads with or without HLA incubated with a control serum without anti-HLA antibodies). The NV represents the level of the anti-HLA antibody. Anti-HLA antibodies generally have broad reactivity against multiple HLA class I and class II antigens. In these analyses, we used the highest anti-HLA-I NV and the highest anti-HLA-II NV to define the antibody level. From the SAFB assay, we also obtained the following parameters: antigen specificity of the antibody and DSA of the antibody.
Histologic data for these analyses were obtained from surveillance and clinical biopsies performed to evaluate allograft dysfunction. Surveillance biopsies form an integral part of the management of kidney transplant recipients at our institution. Typically, biopsies are carried out at the time of transplantation (time 0) and at 4, 12, 24, and 60 months after transplantation. All patients in this study underwent surveillance biopsies at least at the time of transplantation and after 1 year. All biopsies were evaluated by routine light microscopy and scored using the Banff ’97 classification (1). Immunofluorescence and electron microscopic evaluation were not systematically performed in these biopsies. The diagnoses of acute cellular or antibody-mediated rejection (AMR) were based on published criteria (7). The diagnosis of TG in this study was based on the presence of duplication of the glomerular basement membrane (Banff cg score >0) detected by light microscopy using silver-methenamine staining in the absence of evidence of recurrent glomerulonephritis or thrombotic microangiopathy. All biopsies with a cg score greater than 0 were reexamined by one of the authors (S.S.) and rescored. The result of the C4d staining was not considered in the diagnosis of TG. Sixty-eight of the 73 patients (93%) had C4d stains in the biopsy. In 34 cases, the C4d stains were performed by immunofluorescence at the time of the biopsy, and in 34 cases, the C4d stains were performed retrospectively using the immunoperoxidase technique. In five cases, the tissue was not available for C4d staining.
Proportions were compared by chi-square analysis. When data were distributed normally comparisons between groups were performed by Student’s t test for two groups or analysis of variance for more than two groups. For not normally distributed data nonparametric tests were used. To assess the relationships between antibody level and other parameters, we used for each patient the single highest antibody NV against HLA-I and the single highest NV against HLA-II. Graft survival in patients with TG was analyzed from the time of TG diagnosis to the end of the follow-up and censored at the time of patient death. Survival analyses were performed by Cox regression and Kaplan-Maier plots.
Among the 598 patients included in this study, 39% had detectable anti-HLA antibodies before transplantation. Compared with antibody-negative patients (Table 1), recipients with anti-HLA antibodies at transplant were more likely to be women, younger, and recipients of second transplants.
Table 2 displays the characteristics of anti-HLA antibodies. These antibodies had reactivity only against HLA-I or HLA-II, or against both HLA classes of antigens. The incidence of DSA was significantly higher among patients with anti-class II antibodies than among patients with anticlass I antibodies (P=0.0002, chi-square). The NV for antibodies against HLA-I and against HLA-II were not significantly different. However, the patients who had reactivity against both HLA-I and HLA-II had higher NV than the patients who had reactivity only against HLA-I or only against HLA-II, respectively (Table 2). Patients with anti-class I antibodies had a similar incidence of antibodies against A (46%) or B (54%) antigens. Patients with antibodies to class II had similar incidence of antibodies directed against DR (60%) and DQ (40%) antigens.
Anti-Human Leukocyte Antigen Antibody Characteristics and Risk of Transplant Glomerulopathy
As shown in previous studies (2), in T-cell crossmatch-negative recipients, the risk of TG was associated with the presence of anti-HLA II antibodies (hazard ratio [HR]=7.77, CI [5.4–13.2]; P<0.0001) but not with the presence of anti-HLA-I antibodies alone (P=0.303). Furthermore, among patients with anti-HLA-II antibodies, the risk of TG was significantly higher in those with DSA than in those without DSA (HR=2.05 [1.03–4.03]; P=0.041), and the risk of TG was also associated with increasing anti-HLA-II antibody levels (NV), irrespective of whether it had DSA or not. Figure 1 displays this latter relationship. The incidence of TG was 5.3% in patients without anti-HLA-II antibodies pretransplant, 18% in patients with NV <2000, 38% in patients with NV between 2001 and 10,000, and 36% in patients with NV >10,000 (P<0.0001, chi-square). Quantitatively, compared with patients without anti-HLA-II, the risk of TG was increased in patients with NV <2000 (n=40) (2.59 [1.07–6.27]; P=0.034). The risk was significantly higher in patients with NV between 2000 and 10,000 (n=42) (9.69 [5.07–18.51]; P<0.0001) and in those with NV >10,000 (n=39) (8.37 [4.27–16.39]; P<0.0001).
Among the 166 patients with anti-HLA-II antibodies, 59% also had antibodies to HLA-I (Table 2). The risk of TG was not significantly different between the patients only with anti-HLA-II antibodies and those with antibodies against both HLA-I and HLA-II (data not shown). The risk of TG was not statistically different between patients with antibodies against DR or DQ antigens (P=0.08).
By multivariate analysis (Table 3), anti-HLA-II antibody NV, DSA, and a history of AMR were related to the risk of TG. Additional variables added to this model but found to be nonsignificant included HLA mismatches and transplant number. It should be noted that AMR occurred in 8 of the 598 patients in this cohort. However, 60% of the patients with AMR developed TG during follow-up.
Prognosis of Transplant Glomerulopathy and Relationship With Antibody Characteristics
Among the 598 patients studied, 73 (12%) developed TG during a period of follow-up of 53.5±19 months. Fifty-two of the 598 patients (8.7%) lost their graft (not because of death), including 24 of 525 (4.6%) patients without TG and 28 of 73 (38.4%) with TG (P<0.0001). By Kaplan Meier, the median graft survival after the diagnosis of TG was 43±7 months.
Several histologic findings on the TG diagnostic biopsy related with worse graft prognosis including higher cg score (1.915 [1.183–3.099]; P=0.008), the presence of interstitial inflammation (1.632 [1.067–2.496]; P=0.024), and the presence of C4d in PTC (4.478 [2.025–9.905]; P<0.0001). Glomerulitis (g score) was only marginally associated with graft survival (P=0.065). C4d was considered positive in PTC in 16 of 68 biopsies (24%) and C4d positive allografts had significantly poorer prognosis (Fig. 2). Thus, graft failure occurred in 13 of 53 (25%) with C4d-negative TG compared with 12 of 15 (80%) with C4d positive TG (P<0.0001, chi square). Higher anti-HLA-II NV was strongly associated with C4d positivity in PTC (odds ratio 3.216 [1.376–7.517]; P=0.007 by logistic regression). This relationship was such that C4d was positive in 2 of 28 (7%) patients with no anti-HLA-II antibodies or with these antibodies and an NV <2000. In contrast, C4d was positive in 3 of 15 (20%) patients with NV 2000 to 10,000 and in 6 of 12 (50%) patients with NV >10,000 (P=0.006, chi-square). In contrast to these findings, anti-HLA-II NV did not relate significantly to the presence of inflammation or the cg score in the biopsy.
By multivariate analysis, the relationship between C4d and reduced TG graft survival (HR=3.93 [1.37–11.26]; P=0.011) was independent of other variables that also relate to TG survival including lower GFR (HR=0.962 [0.93–0.99]; P=0.033) and more proteinuria (HR=1.306 [1.05–1.63]; P=0.018) at the time of diagnosis.
These analyses represent a step further in our understanding of the variables that relate to the risk of developing TG and graft survival (2). It is remarkable that despite the relatively short interval between our previous analyses of this patient cohort (2), the number of patients with TG has increased from 55 to 73, supporting the contention that the risk of developing TG increases progressively with the posttransplant time in patients with anti-HLA-II antibodies. These analyses showed that the level of the anti-HLA-II antibody relates to the risk of developing TG. The relationship between the antibody level and the TG risk is discontinuous. That is, beyond a certain NV threshold, the risk of TG does not continue to increase. Possibly, this observation relates to the limitations of measuring antibody levels by fluorescence-based methods that do not provide true antibody titration. Previous studies, using SAFB, showed a relationship between anti-HLA antibody level and graft survival (8). These results suggest that this relationship may be explained by the increased incidence of TG in patients with increasing antibody levels.
The relationship between anti-HLA-II antibody levels and TG was independent of AMR and whether the anti-HLA-II antibody had DSA. We confirmed here that approximately 50% of patients with AMR develop TG (2, 9). However, it is important to note that in most of the patients TG is not preceded by an AMR. Thus, in this study only 7% of patients with TG had a previous episode of AMR. It is likely that antibody-mediated allograft injury may manifest itself as acute inflammation (i.e., AMR) or as protracted endothelial inflammation that is clinically subacute and results in TG. Although the presence of anti-HLA-II antibody DSA increases the risk of TG patients without DSA also have an increased risk of TG. This observation likely indicates that our ability to assess DSA is not completely accurate and that some antibodies deemed DSA-negative by SAFB in fact might have had reactivity against the donor graft. This is indeed consistent with the observation of multiple epitope sharing among HLA-II molecules (10). In this study, we had no information about flow crossmatching, and this is a limitation because flow studies might have revealed non-HLA donor reactivity in recipient’s sera that was not detected by SAFB. Flow studies may have also revealed that some of the patients in this study might have low B-cell flow positivity. However, the lack of flow crossmatch data does not change the basic observation made here that the level of the anti-HLA-II antibody, assessed with the single antigen bead assay, quantifies the risk of developing TG.
These analyses showed that the presence of C4d in PTC is associated with reduced graft survival in TG. Furthermore, these results suggest that the absence of C4d may be due in part to the presence of low levels of antibody. These results link antibody level to graft survival by means of C4d deposition. It has been observed previously that C4d is present in PTC in approximately 30% to 50% of cases of TG (5) despite the presence of antiorgan antibodies (2, 4). It remains unclear whether C4d-negative TG represents lack of complement activation or deposition of C4d below detectable levels. These observations are somewhat weakened by the fact that the C4d deposition was assessed here by immunofluorescence or immunoperoxidase. The sensitivity of these two assays differs (11). Thus, possibly some C4d-negative by peroxidase cases may have been positive by fluorescence. In addition, the use of immunoperoxidase in some biopsy only does not allow us to assess the possible prognostic implications of glomerular C4d (12). The relationship between antibody level, C4d, and graft survival is tantalizing, because it helps us understand the pathogenesis of TG and clinically may give us a tool to better predict graft survival. Furthermore, C4d may also become an indicator of effectiveness of therapy.
The relationship between C4d and TG graft survival was independent of other prognostic factors recognized in previous studies, such as graft function and proteinuria (2) and new ones recognized here. Among the latter, the presence of interstitial inflammation relates to poorer prognosis as shown previously in patients with interstitial fibrosis and tubular atrophy (13–15). The relationship between the extent of glomerular basement membrane duplication and prognosis likely relates to the fact that increasing cg scores relates to disease progression (2).
These studies focused on pretransplant antibody levels. The relevance of newly developed anti-HLA-II on TG was assessed previously (2, 4). These results seem to minimize the role of anti-HLA-I antibodies on TG, but this may be the result of patient selection (T-cell crossmatch-negative recipients) than indicative of lack of pathogenic importance of anti-HLA-I antibodies. Transplant glomerulopathy is clearly a major threat to kidney allograft survival. Early diagnosis by surveillance biopsies is likely important to improve prognosis. These results suggest that reductions in anti-HLA-II antibody levels may reduce the risk of TG and improve graft survival.
The authors thank the kidney-pancreas transplant coordinators for their dedication to the care of transplant recipients and their help in the collection of data from these patients. They also thank Ms. Cynthia Handberg for her excellent secretarial assistance.
1. Racusen LC, Solez K, Colvin RB, et al. The Banff 97 working classification of renal allograft pathology. Kidney Int
1999; 55: 713.
2. Gloor JM, Sethi S, Stegall MD, et al. Transplant glomerulopathy
: Subclinical incidence and association with alloantibody. Am J Transplant
2007; 9: 2124.
3. Monga G, Mazzucco G, Novara R, et al. Intertubular capillary changes in kidney allografts: An ultrastructural study in patients with transplant glomerulopathy
. Ultrastruct Pathol
1990; 14: 201.
4. Sis B, Campbell PM, Mueller T, et al. Transplant glomerulopathy
, late antibody-mediated rejection and the ABCD tetrad in kidney allograft biopsies for cause. Am J Transplant
2007; 7: 1743.
5. Colvin RB. Antibody-mediated renal allograft rejection: Diagnosis and pathogenesis. J Am Soc Nephrol
2007; 18: 1046.
6. Cosio FG, Gloor JM, Sethi S, et al. Transplant glomerulopathy
. Am J Transplant
2008; 8: 492.
7. Racusen LC, Colvin RB, Solez K, et al. Antibody-mediated rejection criteria—An addition to the Banff 97 classification of renal allograft rejection. Am J Transplant
2003; 3: 708.
8. Mizutani K, Terasaki P, Hamdani E, et al. The importance of anti-HLA-specific antibody strength in monitoring kidney transplant patients. Am J Transplant
2007; 7: 1027.
9. Gloor JM, Cosio FG, Rea DJ, et al. Histologic findings one year after positive crossmatch or ABO blood group incompatible living donor kidney transplantation. Am J Transplant
2006; 6: 1841.
10. Cai J, Terasaki PI, Mao Q, et al. Development of nondonor-specific HLA-DR antibodies in allograft recipients is associated with shared epitopes with mismatched donor DR antigens. Am J Transplant
2006; 6: 2947.
11. Nadasdy GM, Bott C, Cowden D, et al. Comparative study for the detection of peritubular capillary C4d
deposition in human renal allografts using different methodologies. Hum Pathol
2005; 36: 1178.
12. Sijpkens YW, Joosten SA, Wong MC, et al. Immunologic risk factors and glomerular C4d
deposits in chronic transplant glomerulopathy
. Kidney Int
2004; 65: 2409.
13. Shishido S, Asanuma H, Nakai H, et al. The impact of repeated subclinical acute rejection on the progression of chronic allograft nephropathy. J Am Soc Nephrol
2003; 14: 1046.
14. Cosio FG, Grande JP, Wadei H, et al. Predicting subsequent decline in kidney allograft function from early surveillance biopsies. Am J Transplant
2005; 5: 2464.
15. Moreso F, Ibernon M, Goma M, et al. Subclinical rejection associated with chronic allograft nephropathy in protocol biopsies as a risk factor for late graft loss. Am J Transplant
2006; 6: 747.